Lark CLI Skill: Skill Maker — Create Custom Skills

What it is

The Lark CLI Skill Maker is a skill for creating reusable custom skills on the Lark (Feishu) platform. It enables developers to wrap atomic API calls or orchestrate multi-step workflows into shareable skill definitions that other users and AI agents can invoke.

The skill targets teams building automation on Lark/Feishu who want to encapsulate common operations (sending messages, creating documents, managing calendars) into reusable, composable building blocks.

How it saves time or tokens

Instead of writing the same Lark API integration code for every project, the Skill Maker packages it into a reusable skill definition. Other developers or AI agents can invoke the skill by name without understanding the underlying API details.

The composable skill architecture means complex multi-step workflows (like 'create a project channel, invite members, and post a kickoff message') become single skill invocations.

Additionally, the project's well-structured documentation and active community mean developers spend less time troubleshooting integration issues. When AI coding assistants generate code for this tool, they can reference established patterns from the documentation, producing correct implementations with fewer iterations and lower token costs.

How to use

1. Define a skill by specifying the API calls and parameters:

name: create-project-channel
description: Create a Lark group chat for a new project
parameters:
  - name: project_name
    type: string
    required: true
  - name: member_ids
    type: array
    required: true
steps:
  - action: create_group_chat
    params:
      name: '{{project_name}} Channel'
  - action: add_members
    params:
      chat_id: '{{steps[0].result.chat_id}}'
      members: '{{member_ids}}'

1. Register the skill with the Lark CLI:

lark-cli skill register ./create-project-channel.yaml

1. Invoke the skill:

lark-cli skill run create-project-channel \
  --project_name 'Q2 Launch' \
  --member_ids '["user1", "user2"]'

1. Share the skill with your team for reuse.

Example

# Programmatic skill invocation
from lark_cli import SkillRunner

runner = SkillRunner()
result = runner.run('create-project-channel', {
    'project_name': 'Q2 Launch',
    'member_ids': ['user1', 'user2']
})
print(f'Channel created: {result["chat_id"]}')

Related on TokRepo

• AI Tools for Automation — Workflow automation tools and platforms
• AI Tools for Task Management — Task and project management tools

Common pitfalls

• Creating skills that are too specific. Design skills with configurable parameters so they can be reused across different projects and contexts.
• Not handling API rate limits in multi-step skills. Lark APIs have rate limits. Add delays or retry logic between steps to avoid throttling.
• Skipping error handling in skill definitions. Multi-step skills should define rollback behavior when intermediate steps fail.
• Failing to review community discussions and changelogs before upgrading. Breaking changes in major versions can disrupt existing workflows. Pin versions in production and test upgrades in staging first.